Are you interested in working with real-time data? If so, join us in VORN’s challenge to provide spectators with insights based on measurements from smart shirts!
Biathletes must be able to control their heart and breathing rates while shooting. This effort is something that spectators can't see. The goal of the challenge is to create a proof of concept for accessing, transforming and visualizing the real-time measurements of the athletes and making them available to the spectators.
For this challenge, you will be working with Kafka. The development environment, real-time data simulation and streaming is set up for you. That’s where your journey begins!
Help us with your knowledge, skills and creativity to further process the data of individual athletes. Which metrics will we need? How should they be aggregated and queried? What is the best way to visualize them to the spectators? Can we use audio signals or video overlays?
VORN has created a smart shirt that can measure a range of physical and mental metrics.
We implemented a real-time streaming architecture via Apache Kafka with which VORN can access the data ongoingly. We then visualized the data on Tipboard. Using rules of thumb, we created a new variable for shooting accuracy with three levels: good, medium and bad. On the Tipboard dashboard, we also displayed the beats per minute (BPM), besides HRV and predicted shooting accuracy, to show the athletes’ stress level and ability to lower their breathing rate in preparation for a shot. Visualizing such factors helps TV viewers to better understand the variables influencing the shooting score.
Potential next steps
- Improve the data quality (reduce the number of null values)
- Find a way to measure stress
- Collect more data on HRV and BPM to make a benchmark for each individual athlete. It will help to make the shooting accuracy prediction more personalized to each individual athlete.
- Record shooting score and correlate with the available data.
VORN is still developing the product, and only the first test data is available. However, these sets make it possible to simulate the data stream in real-time. The data is collected as physical data, which is then transformed into indicators such as heart rate. The data has already been cleaned from noise.
Each measurement comes with a timestamp and other attributes such as the core ID of the shirt, which can be assigned to an individual athlete.